A Discussion About Ratings Compression
June 4, 2014
There has been a perception that Arbitron’s (now Nielsen) transition to PPM in the country’s largest markets has resulted in ratings compression. It is a hard idea to refute when you look at an AQH rating ranker and find multiple stations tied with a 0.5 (for example), and often only 0.1 or 0.2 of a rating point can separate a dozen radio stations in a given market. If you accept the notion that perception is reality, then all of these stations are tied, and on a cost per point basis there is no difference.
We at Research Director, Inc. do not intend to argue whether or not ratings compression exists. Instead, we would like start a conversation about how it affects our industry. As broadcasters, we tell advertisers that they should ignore CPP and instead buy the value of the listeners by focusing on qualitative factors. It’s a great strategy (and should always be a part of the story), but not always realistic for the busy media buyer whose primary goal is to bring in the buy at their designated CPP. This is an unfortunate reality in our world.
Before we continue, let’s review the calculation that is used to determine GRPs. It is simply the number of spots in the schedule multiplied by the reported AQH rating. Take those stations that are tied at the top of the rating ranker, and you get identical GRPs. Compression at work.
But the reality is that there may be a large quantitative difference between stations with the same reported rating because ratings are rounded to one decimal place. As an example, a rating of 0.3 could be anywhere from 0.2500 to 0.3499 – that’s a 40% difference! One possible solution would be to report ratings to more decimal places, therefore showing that quantitative difference.
There is another option. Presently GRPs are calculated by multiplying the number of spots in the schedule by the station’s AQH rating. For example, 10 spots on a station with a 0.5 rating would yield 5.0 GRPs. What if, instead of using the station’s AQH rating, we calculated a station’s gross impressions (number of spots multiplied by AQH persons) and then divide by the population? Using this method, two stations with a difference of just 100 AQH persons will yield different GRPs. When you compare stations’ AQH persons instead of their rounded ratings, there will be far fewer ties.
In our opinion, and while not the only reason, the perceived compression in PPM markets has hurt broadcasters’ pricing and encourages a reverse auction, where the lowest CPP wins. By calculating CPP more precisely, this downward pricing pressure may be diminished.
With all this said, we would be remiss if we did not state a couple of key points – that is, the downside of changing the industry-accepted method of calculating GRPs and CPP.
The first point is reliability. Reliability increases with larger sample sizes and by making the original sample more representative of the population that it is measuring. Reporting the AQH rating to two or more decimal places might be more precise, but it does not improve the reliability of the estimate. The statistical reliability of that estimate, expressed as a margin of error around that number, is exactly the same no matter how many decimal places are shown. So our proposal to use AQH persons to calculate GRPs and CPP and round to more decimal places will be more precise, but not more accurate or reliable.
The second point is simple math. If you use an unrounded rating, on average, half the time it will be lower and half the time it will be higher. So you can charge more half the time, but less the other half. However, if buyers can see the difference between stations that are tied based on rating, we still might be able to avoid the reverse auction, and that will benefit all broadcasters.
We would like to continue this discussion, and welcome your opinion. Tweet your thoughts and include #ratingscompression and @ResearchDirectr, or feel free to comment below.
-Charlie Sislen, Partner